Hi,

Thanks for the amazing work on Sora2-mini / UniAVGen, the demo looks fantastic, and this unified audio-video generation framework is very inspiring!

In the demo page, there are 5 multi-task capabilities shown:
1. Joint Audio-Video Generation
2. Joint Generation with Reference Audio
3. Joint Audio-Video Continuation
4. Video-to-Audio Dubbing
5. Audio-Driven Video Synthesis

However, in the Sora2-mini repository, here is no direct support or example for “Joint Audio-Video Continuation”. The task where conditioned audio/video input is continued seamlessly, matching style and timing of the conditional content, even though it’s described on the demo site.

Could you please clarify whether Joint Audio-Video Continuation is currently supported or are there any plans to add it in the future?

Thanks again for this excellent project!

Best,
Georgi